Atopy patch test

Compared with the DBPCFC test, the atopy patch test (erythema usually with infiltration) had high specificity (81–96%) for any reaction (immediate, delayed, or immediate and delayed combined) to cow’s milk, egg and soya. Specificity for any reaction to wheat was more variable (35–94%). Compared with an open food challenge, the specificity results for any reaction were more variable for cow’s milk, egg, wheat and peanut. Compared with DBPCFC or an open food challenge, sensitivity results for any reaction to a single food (cow’s milk, egg, wheat, soya and peanut) were more variable. Sensitivity and specificity results compared with DBPCFC were both more variable when considered for several foods together (no data compared with an open food challenge).

Skin prick test

Compared with the DBPCFC test the skin prick test (wheal size 3 mm or greater) had high sensitivity (90–95%) for diagnosing an immediate response to fish and peanut, or to several foods together (results from one study); specificity results for these foods were more variable. The sensitivity and specificity for detecting any reaction (immediate, delayed, or combined and compared with DBPCFC) to all other allergens tested (cow’s milk, wheat and soya) were more wide-ranging across studies. Compared with an open food challenge, sensitivity and specificity results for any reaction (immediate, delayed, or combined) to all allergens were also more variable.

Specific IgE

The sensitivity of specific IgE (more than 0.35 ku/l) for detecting any reaction (immediate, delayed, or combined) to cow’s milk and egg was high (83–100%) compared with the DBPCFC test. Sensitivity for detecting an immediate reaction to wheat, soya, fish and peanut compared with DBPCFC was also high (94–97%; one study only). Sensitivity for a combined immediate and delayed reaction to wheat or soya was more variable (no data for delayed reactions). Specificity results for each of the allergens alone or when considered together were more variable. Compared with an open food challenge, both sensitivity and specificity results were less consistent across all foods tested. The specific IgE level indicative of a positive test ranged from 0.35–99 ku/l in the open challenge studies.

Effect of changing test parameters

The available data for each type of test do not show consistency in sensitivity or specificity results. This might reflect the way the particular tests were undertaken or the criteria used to define positive test results. Several studies have considered whether changing certain parameters of a test affects their diagnostic accuracy in children with atopic eczema.^{165,167,175,182–191} The accuracy of the atopy patch test varied according to the size of the chamber used for occlusion, the vehicle and concentration used to apply the allergen to the skin, and according to which skin sign was taken to indicate a positive test. [EL = 3/EL = DS III] There was some evidence that increasing the wheal size that constituted a positive test on skin prick testing increased the specificity of the test. The specific IgE levels that gave PPVs of 95% for certain allergens were estimated in one study. [EL = DS III]

Outcome of challenge tests

One study recruited 63 children (age range 0.4–19.4 years) with persistent atopic eczema from a dermatology clinic. After screening for food allergy, 19 children underwent DBPCFC. There were four positive challenges to egg, three to milk, three to wheat, two to barley and one to beef. Thirty-seven percent of children in this study were diagnosed with food allergy based on double-blind and open challenges and convincing histories of reactions to foods.¹⁴³

Three studies conducted in a single clinic reviewed patient records of consecutive referrals for food allergy investigation in children with atopic eczema. In the first study, 107 children (age range 5 months to 12 years, median age 21 months) underwent DBPCFC. Eighty-seven children (81%) showed reactions to at least one food. Early skin reactions were mainly caused by egg and cow’s milk, with wheat and soya producing a small number of reactions. Isolated late reactions were produced by all four allergens in 25% of the total number of reactions.¹⁶⁶

The second study reviewed 139 children (age range 2 months to 11.2 years, median age 13 months). DBPCFC was conducted in all children and, of 208 challenges, 111 were positive. Positive challenges were due to cow’s milk (47%), egg (34%) and wheat (19%). It is not clear how many children had at least one positive test. Immediate and late reactions were observed.⁵⁴⁷

There were 98 children in the third study (age range 2 months to 11.2 years, median age 13 months). Of 173 challenges, 55% were positive, including 45/71 positive challenges to cow’s milk, 28/42 to egg, 18/35 to wheat and 4/25 to soya. It is not clear how many children had at least one positive result. All late reactions (25% of the total) were skin related.¹⁶⁵

In one small study, 26 children (age range 16 months to 19 years, median age 11 years) with atopic eczema, elevated IgE level), suspected food allergy and ability to cooperate with food challenge procedures were subjected to DBPCFC. Fifteen children had positive reactions to at least one food. Foods provoking cutaneous reactions were wheat, soya, milk, egg, rye, chocolate and chicken.⁵⁴⁸

A later study by the same group investigated 113 people (age range 4 months to 24.5 years, median age 6 years) with severe atopic eczema for food allergy. Three hundred and seventy DBPCFCs were undertaken, of which 101 were positive in 56% of patients. Major allergens in this study were egg and peanut, with 85 challenges provoking skin reactions. Allergens provoking a small number (fewer than 6) of reactions were milk, soya, wheat, fish, chicken, pork, beef and potato.⁵⁴⁹

In another study, food allergy was identified by DBPCFC in 39% (n = 165) of patients (age range 4 months to 21.9 years, mean age 49 months) with atopic eczema attending a specialist allergy clinic. The main foods provoking positive challenges within 2 hours were peanut (n = 27), egg (n = 33) and milk (n = 14). No delayed reactions were observed.⁵⁵⁰

Sixty-four children (age range 1–10 years, median age 2 years) with atopic eczema and suspected food allergy were investigated in this study. There were 106 DBPCFCs, of which 46% were positive. The allergens tested were cow’s milk, egg, wheat and soya. All of the isolated late reactions (12% of total) were eczematous.¹⁶⁹

Seventy-four children (age range 6 months to 16.3 years, median age 2.5 years) with atopic eczema who had been referred to paediatric dermatology or allergy clinics were enrolled for evaluation of food allergy. Six children underwent DBPCFC and positive reactions were to milk (three challenges) and wheat (two challenges).¹⁴²

Cow’s milk and egg exclusion diets

Two double-blind randomised crossover trials of egg and cow’s milk exclusion diets involved children with atopic eczema.^193,194 [EL = 1-] The studies had 4 or 6 week treatment periods, with a washout period of the same duration in-between. As well as eliminating eggs and cow’s milk, chicken and beef were eliminated, and a soya-based milk substitute given; the control group received a preparation containing a mixture of dried eggs and cow’s milk as a milk substitute. Neither study stated whether there was clear evidence of allergy or intolerance to the eliminated foods, although it was reported in one that three of the 20 children who completed the study had a history of exacerbation of skin symptoms following ingestion of eggs or cow’s milk.¹⁹³ The most common reason for withdrawal from both studies was non-adherence to the diet. Both studies analysed results only for those who completed treatment.

The first RCT (n = 36; 56% completed), in children aged 2–8 years, found significantly greater improvements in the diet group versus control in atopic eczema activity (global improvement) and skin area affected, sleeplessness and antihistamine usage, with no significant difference between diet and control groups in pruritus (mean improvement 4.49, standard error 2.25).¹⁹³ The response in the diet group was significantly greater during the first treatment period than the second treatment period for activity, area and sleeplessness, but there was no significant difference between the first and second treatment periods for pruritus or antihistamine usage. For pruritus and sleeplessness this ‘order’ effect was greater than the difference between diet and control groups. It was also reported that there was no correlation between positive prick test to the egg and cow’s milk antigens and response to diet, but no data were reported.¹⁹³

The second RCT, in children and adults aged 1–23 years, found no significant differences between elimination and control diets in area (mean −1, 95% CI −6 to 3.4) or itch (mean 15, 95% CI −21 to 51) scores. Use of topical corticosteroids was higher during the elimination diet (n = 53; 40 completed).¹⁹⁴

Two case series also reported the effects of egg and/or cow’s milk exclusion diets in children with atopic eczema.^195,196 One series eliminated cow’s milk and egg from the diet of children (aged 0.4–15 years) for 3 weeks (n = 91; 73% completed and analysed). Improvements in severity scores were reported at endpoint.¹⁹⁶ In the other series, children aged 2–14 years who had not responded to usual treatments eliminated cow’s milk and egg, or cow’s milk only, from their diet for 4 weeks. The decision on whether to exclude milk alone or both foods was dependent on which was suspected of precipitating the atopic eczema. However, the outcome was only reported as cure or improvement, with no definition of either term. Additionally it was not clear how many of the children eliminated only milk or both foods from their diet (n = 59).¹⁹⁵ [EL = 3]

One case series of children with atopic eczema (n = 11, median age 4 years) documented acute allergic reactions to cow’s milk after prolonged cow’s milk elimination diets.¹⁹⁷ [EL = 3]

Egg exclusion diets

Two controlled trials considered the effects of egg exclusion on atopic eczema in infants.^198,199 The first was a double-blind RCT in which all the infants had a raised IgE to egg on a radio-allergosorbent test (RAST), and the majority also had a positive test on a DBPCFC test (n = 62; 89% analysed).¹⁹⁸ The control group were not given any specific dietary advice. After 4 weeks’ intervention, the reduction in body surface area affected was significantly greater in the diet group compared with control (mean difference 5.25%, 95% CI 0.1% to 10.9%, P = 0.04). Differences between groups in severity scores were not significant (6.1, 95% CI −0.1 to 12.3, P = 0.05), although there were some discrepancies in the trial report between data presented in the text and in the abstract.¹⁹⁸ [EL = 1−]

The second trial, described as a single-blind controlled study, reported the proportions of children in four age categories whose condition was ‘better’ after 2 weeks’ treatment. However, ‘better’ was not defined (n = 213; 65% of whom completed treatment and were analysed).This study was not considered further.¹⁹⁹ [EL = 1−]

Cow’s milk substitutes

One RCT compared two milk substitutes in infants with atopic eczema and allergy to cow’s milk (shown on DBPCFC; n = 73).²⁰⁰ An amino acid-based formula was compared with a hydrolysed whey formula. Energy intake was similar in both groups. A significant improvement in the SCORAD severity index was seen overall, from a mean of 24.6 at entry to 10.7 after 6 months (P < 0.0001); data were not reported separately by treatment group. In the amino acid group there was a significant increase in the length standard deviation score (SDS) from baseline (P < 0.04), while there was no statistically significant change in the hydrolysed whey group. Weight-for-length values were ‘stable’ in both groups.²⁰⁰ [EL = 1−]

In a randomised study infants with atopic eczema and proven allergy to cow’s milk (on double-blind food challenge) were given hydrolysed whey or amino acid formulas as milk substitutes (n = 45).²⁰¹ Although the study was described as randomised in the abstract, randomisation was not mentioned elsewhere in the paper. Other dietary restrictions (egg and cereals) were also used in two-thirds of infants. At 8 months, SCORAD scores had improved significantly from baseline in those receiving either milk substitute. The statistical significance of changes in weight and length of infants was also reported, although the data were only presented in graphs. The graphs showed that weight and length increased in both groups in the first month of treatment, and they continued to increase in the amino acid group over the 9 month follow-up period. The pattern in the whey substitute group was less consistent, but weight and length at 9 months appeared to be the same or worse than at baseline. There was overlap of the 95% CIs for the groups for weight indicating that the difference between the groups was not statistically significant for this outcome. However, the difference between groups for length was statistically significant.²⁰¹ [EL = 1−]

Milk substitutes for women who are breastfeeding

One double-blind crossover RCT considered the effects of an exclusion diet plus a milk substitute in mothers of breastfed infants with atopic eczema (n = 19; 17 completed and analysed; aged 6 weeks to 6 months).²⁰² [EL = 1−] This was the only study relevant to the guideline clinical question in a review of maternal dietary antigen avoidance during pregnancy and/or lactation.²⁰³ The foods excluded from the mothers’ diet were cow’s milk, egg, chocolate, wheat, nuts, fish, beef, chicken, citrus fruits, colourings and preservatives. The milk substitutes taken were a preparation containing soya and one containing cow’s milk and egg powder. Area and activity scores (the latter a measure of the severity of the condition on 20 body surface areas) fell from baseline with both milk substitutes after 4 weeks’ use. The difference between groups was not statistically significant (activity score, exclusion diet plus soya versus exclusion diet plus cow’s milk and egg, 104 versus 12.6, P value not reported; area score, exclusion diet plus soya versus exclusion diet plus cow’s milk and egg, 9.0 versus 8.9, P value not reported). A subsequent open, uncontrolled study was undertaken in the same group because of concerns that the soya preparation may have triggered symptoms in the first study (n = 18). In this open study, mothers took their usual diet (containing cow’s milk and egg) for 2 weeks, followed by an exclusion diet for 2 weeks (the same as that taken during the first study), then the usual diet repeated for 2 weeks. Activity and area scores fell significantly after the exclusion diet (at week 4), and remained at around this level after the reintroduction of the usual diet (week 6).²⁰²

Restrictive diets

Three studies considered the effects of restricting foods consumed by children with atopic eczema (two case series and one controlled study).^204–206 The first case series considered a 2 week diet consisting of up to 19 foods (including meats, carrots, lettuce, parsley, pears, rice, plain flour, sugar, golden syrup, honey, oils, vinegar, salt and pepper and coffee; n = 29, age range 2–12 years).²⁰⁴ The withdrawal rate was 55%, and half of the withdrawals were because the diet was considered to be too restrictive. Thirteen children were evaluated at the end of the 2 week diet. According to the parents’ global assessment seven were improved, three remained the same and three deteriorated. Based on the dermatologist’s assessment of inflammation, lichenification, and cracking, five were improved, seven remained the same and one deteriorated.²⁰⁴ [EL = 3]

The second case series included children aged 10 months to 4 years with severe atopic eczema that persisted despite usual treatment and elimination of the food items to which the child was suspected to be allergic (n = 13).²⁰⁶ A diet consisting of the following foods was taken for 1 month: casein hydrolysate, lamb, rice, corn, corn oil, potato, cucumber, melon, bilberries, salt, sugar, and gluten- and milk-free bread. The numbers of children whose condition improved according to investigator’s and parents’ scores of the severity of the condition were six and eight, respectively. Not all the children who improved according to the investigator improved according to the parents.²⁰⁶ [EL = 3]

The controlled study reported changes in IgE and peripheral blood mononuclear cell concentrations following elimination diets (eliminating the ‘offending foods’) in children aged 3 months to 13 years (n = 153). Changes in severity from baseline were also reported, but a lack of between-group analysis and of details of the diets given made interpretation difficult.²⁰⁵ [EL = 2−]

Few foods diets

Three studies considered the effectiveness of ‘few foods’ diets (eliminating all but five to eight foods); these were a single-blind RCT and two case series.^207–211 The single-blind RCT evaluated a diet (including either whey or casein hydrolysate milk formula) in children aged 0.3–13 years with atopic eczema that persisted despite conventional treatment (n = 85).²⁰⁷ After 6 weeks there were no significant differences between the diet group and control group (continued usual diet) in changes in any outcome (body surface area affected, severity, daytime itch, or sleep disturbance). The withdrawal rates were 59% in the diet group (the most common reason for withdrawal being non-adherence), and 15% in the control group; only results for those who completed the 6 week intervention period were analysed.²⁰⁷ [EL = 1−]

A case series of children with extensive atopic eczema (affecting 30% or more of body surface area) that responded poorly to conventional treatment or who had a history of food intolerance were given a few (six) foods diet (n = 63, age range 0.4–14.8 years).^208,210 After 6 weeks, the median severity score fell by 33%, with 52% having a 20% or greater reduction in score. ‘Little or no benefit’ was seen in 39%. The withdrawal rate was 14%. Of the 68% who were followed up for 1 year, the outcome was similar in children regardless of their response to the 6 week few foods diet, although no data were presented.²⁰⁸ Some children from this study were subsequently given an elemental diet (see below).^209,210 [EL = 3]

Another case series of children with severe atopic eczema evaluated a few foods diet (n = 66, age range 0.6–17 years).²¹¹ Twenty-four patients (36%) were reported to have ‘worthwhile’ improvement (the term ‘worthwhile’ was not defined) from the diet (median duration 26 days, range 19–44 days). In 15 of these (23% of the total group), improvement persisted on dietary treatment, but three withdrew because the diet was too burdensome. Overall, 12 (18%) persisted with the diet and had continued benefit over the duration of follow-up (mean 48 weeks, range 26–71 weeks). The outcomes beyond this follow-up period were not reported.²¹¹ [EL = 3]

Elemental diets

A randomised crossover study in infants and children with a positive skin prick test and raised cow’s milk-specific and soya bean-specific IgE evaluated an amino acid-based elemental diet (n = 15; 11 analysed).²¹² Dairy- or soya-based products were also excluded from the diet. The control group continued with a pre-existing formula (no further details were reported). Following 6 weeks’ treatment, there were no significant differences between the amino acid-based elemental diet and the control diet in SCORAD scores or in global health scores. A significant treatment-by-period interaction was reported for SCORAD, which was greater than the between-group treatment difference.²¹² [EL = 1−]

A case series of children aged 0.4–13 years with severe and extensive atopic eczema were hospitalised for treatment with an elemental formula only (the product contained 100% free amino acids). Pet and house dust mite avoidance measures in the children’s homes were a prerequisite for treatment (n = 37).^209,210 After a median duration of 30 days’ treatment, 27% were considered to be ‘treatment failures’ because their severity scores were unchanged or worse compared with baseline. In the 73% for whom treatment was considered to be successful, the severity scores decreased to 27% of the baseline score (range 3–67%; no further details of who had greatest or least benefit were presented). No significant differences in demographics or in clinical features were found between those in whom treatment was successful and those in whom it was not successful. Reported adverse effects were weight loss of up to 17% (in 30 of 34 evaluated), loose stools (19%), and a reduction in serum albumin in 93% of 27 children in whom this parameter was measured (from a mean of 30.8 g/l to a mean nadir of 21.2 g/l). No electrolyte disturbances were reported.^209,210 [EL = 3]

A further case series reported the outcomes of an elemental diet in children with atopic eczema (n = 10, age not specified).²¹³ Only the elemental diet was used for 2 weeks, followed by addition of pumpkin, potatoes, zucchini, apples, pears and pure vegetable margarine. Two children stopped using the diet after 1 week. In the other eight, the atopic eczema scores (a measure of severity, extent and of treatment required) fell significantly at 6 weeks, and increased again after reintroduction of their usual diet. Adverse effects were not considered. [EL = 3]

The effectiveness of a ‘home-made meat-based formula’ diet was considered in a case series of children with severe atopic eczema (n = 16, aged 5–24 months).²¹⁴ The children had positive skin prick test results to cow’s milk, egg, and wheat and/or soya. The formula consisted of lamb, olive oil, rice flour and water, supplemented with calcium and vitamin D. After 1 month, the severity score had fallen (no statistical analysis reported), with no significant changes in lipid levels. It was reported that all the children had gained weight normally, but no data were presented.²¹⁴ [EL = 3]

Sugar exclusion

One study considered whether avoiding sugar had an impact on atopic eczema in children and adults (n = 30; 9 children).²¹⁵ No significant changes in SCORAD severity scores were seen in the children’s atopic eczema 1 week after the elimination diet, and differences in SCORAD following a double-blind or placebo food challenge were also not significant. Aspartame was offered as a replacement for sugar, but it was not clear how many took this.²¹⁵ [EL = 3]

Sodium cromoglicate

Four studies evaluated the effectiveness of sodium cromoglicate therapy in children with atopic eczema, either compared with or in addition to an elimination diet (three RCTs and one case series).^216–219 The first RCT compared a restricted diet (consisting of 12 foods) with oral sodium cromoglicate in children aged 5 months to 14 years. After 4 weeks’ treatment, there were no significant differences between groups in severity or disease extent (n = 1085, 80% analysed).²¹⁶ [EL = 1−]

Two placebo-controlled crossover RCTs evaluated the addition of sodium cromoglicate to an elimination diet tailored to individual children with atopic eczema.^217,219 In the first RCT, significant improvements in severity were reported for both groups after 6 weeks’ treatment, but no between-group analysis was reported to allow comparison between groups (n = 29, 76% completed and analysed, aged 3–12 years).²¹⁷ [EL = 1−]

The second crossover RCT found no significant differences between investigators’ or parents’ assessments of severity at 8 weeks when treatment with sodium cromoglicate was followed by placebo. However, improvements in severity were significantly greater with sodium cromoglicate when the treatment sequence was reversed (that is, placebo taken first, n = 31, 94% analysed, aged 6 months to 10 years).²¹⁹ [EL = 1−]

In the case series, sodium cromoglicate was added to an individually tailored exclusion diet in children aged 1–15 years (n = 35). However, the outcome of sodium cromoglicate treatment was only expressed as ‘improved’ or clear/almost clear, with no definitions given. Without a control group the study was of limited value, and it was not considered further.²¹⁸

Vitamin and mineral supplementation

Two placebo-controlled RCTs considered the effectiveness of zinc or vitamin E for atopic eczema.^220,221 The trial involving zinc included children aged 1–16 years who continued with their usual treatments for atopic eczema (emollients and topical corticosteroids). Itch scores were significantly higher in children treated with zinc than with placebo, otherwise there were no significant differences in any outcome at 8 weeks (sleep disturbance, redness, surface area or combined disease severity scores, or in use of other treatments; n = 50, 84% analysed).²²⁰ [EL = 1−]

The trial of vitamin E included children and adults (n = 96, aged 10–60 years).²²¹ Treatment with emollients was continued. Vitamin E or placebo was given for 8 months, after which the global assessment of the condition (classifications not defined) found worsening in 8% of the vitamin E group versus 78% in the placebo group; no change in 12% versus 11%, slight improvement in 20% versus 9%, great improvement in 46% versus 2%, and almost complete remission in 14% versus 0%. No statistical analysis of the data was presented and no adverse effects were reported.²²¹ [EL = 1−]

Probiotics

Three double-blind RCTs considered the effectiveness of a milk substitute supplemented with probiotics for the treatment of atopic eczema in infants with suspected cow’s milk allergy.^222–224 The cow’s milk substitute in all three studies was a hydrolysed whey formula, with Lactobacillus added in the intervention group. Two studies had an additional intervention group: one received a mixture of probiotics (Lactobacillus, Bifidobacterium, Propionibacterium) and the other received Lactobacillus rhamnosus. Control groups received the hydrolysed whey formula only. Two studies evaluated 1 month’s use. Of these, the study with three treatment arms found no significant differences between the groups treated with probiotics and the control group in changes in SCORAD severity scores (n = 252, 91% completed and analysed).²²² [EL = 1−] The second study reported significant improvements in SCORAD scores from baseline in the group receiving the hydrolysate plus probiotic. However, no between-group analysis was reported (n = 31).²²³ [EL = 1−] The treatment period in the remaining study was 3 months and no differences in SCORAD reduction were found between the three groups.²²⁴ [EL = 1−]

House dust mite avoidance

Two RCTs in children^225,226 and one involving children and adults²²⁷ considered the effectiveness of house dust mite avoidance. One of the RCTs evaluated the effects of bedding encasement with microfine fibres on mite sensitisation in children with atopic eczema, but did not report any clinical outcomes (only IgE and house dust mite levels were measured) and is therefore not considered further (n = 57).²²⁵

A 2 month placebo-controlled RCT in young children (aged 2–10 years, mean 3.9 years) evaluated house dust mite allergen avoidance measures. The children had moderate atopic eczema (SCORAD 27–33) associated with high total and/or specific IgE serum levels (n = 41).²²⁶ The mite avoidance measures consisted of encasing mattresses and pillows, a hot weekly wash of bedding, vacuuming of living rooms and bedrooms at least twice a week, and removing or washing soft toys once a week; pets were not allowed. In the control group the previous house cleaning strategy was continued. After 2 months’ intervention, a significant reduction in the SCORAD index was reported in the avoidance group; the score also fell in the control group, but no between-group analysis was reported. Significant reductions from baseline in dust load and house dust mite allergen concentrations were reported in the avoidance group, but not in the control group; again no between-group analysis was reported, nor was there any consideration of whether groups were similar at baseline in the parameters measured.²²⁶ [EL = 1−]

A further double-blind RCT (n = 60, aged 7–65 years) compared a house dust mite avoidance strategy (GORE-TEX® bedding system, carpet spraying and use of a high-filtration vacuum cleaner) with placebo in children and adults who had positive results in skin prick tests using a range of inhalant allergens. ²²⁷ After 6 months, the reduction in severity (measured using SASSAD) was significantly greater in the avoidance group compared with placebo (mean difference 4.2, 95% CI 1.7 to 6.7, P = 0.008; mean difference in final severity score in children (aged younger than 17 years) 11.1, 95% CI −3.1 to 25.3, P = 0.019). The reduction in bed mattress dust load was significantly greater in the intervention group compared with placebo (98% versus 16%, P = 0.002). Median reductions in the concentrations of the house dust mite allergen in bedroom or living room carpets were not significantly different between intervention and control groups (91% versus 89%, P = 0.94 and 76% versus 38%, P = 0.27, respectively).²²⁷ [EL = 1−]

In a non-randomised controlled study, the effectiveness of an air cleaning system (in a ‘clean-room’) for the treatment of people aged 8–75 years with atopic eczema who had high specific IgE levels to house dust mite was evaluated (n = 30).²²⁸ Participants were hospitalised for 3–4 weeks, and were exposed to either an air cleaning system in a clean-room, or to a similar room without the air cleaning system. The only clinical outcome reported was time to recurrence of symptoms – it was unclear whether this referred to all symptoms or specifically to itchiness. It was reported that time to recurrence of symptoms in those in the clean room who had high IgE to house dust mite was a mean of 8.4 months, whereas in those with no raised IgE to house dust mite the time to recurrence was 1.7 months. In the control group (no air filtration system, and high IgE to house dust mite) the time to recurrence was 1.6 months. No baseline data were reported.²²⁸ [EL = 2−]

Hyposensitisation to house dust mite

Two studies considered the effects of hyposensitisation to house dust mite on atopic eczema in children who had a positive skin prick test result to this allergen. One was a double-blind RCT with 6 months’ follow-up (n = 26),²²⁹ [EL = 1−], and the second was a controlled trial of up to 3 years’ duration (n = 60).²³⁰ [EL = 2−] Neither study found significant differences in the severity or clinical features of atopic eczema between those receiving hyposensitisation therapy and those in the control groups (placebo or continued usual treatment).^229,230

Potential trigger factors

A plethora of potential triggering factors for atopic eczema has been documented in the scientific literature, including irritants, contact allergens, food and dietary factors, inhalant allergens, microbial colonisation of skin, climate, environmental factors and familial factors. Many of these have been considered only in the context of primary causes/prevention of atopic eczema (which are outside the scope of this guideline), rather than in terms of triggering exacerbations of established atopic eczema. Most data in relation to the identification and management of trigger factors relate to testing for food allergies and elimination diets, and avoidance strategies for inhalant allergens.

Identification of trigger factors

There has been little consistency among the studies that have considered the accuracy of atopy patch tests, skin prick tests and specific IgE for identifying food allergy in children with atopic eczema. The studies varied in the age of the study populations, the foods tested, the standard against which results were compared (DBPCFC or open food challenge), and in the way the tests were undertaken (the types of foods used and the criteria used to define positive test results). There was evidence that changing the definition of a positive test result for the atopy patch test, the skin prick test and specific IgE changed the diagnostic accuracy of the tests. [EL = DS III]

Only a minority of studies focused on delayed reactions (in which the suspected food caused exacerbation of atopic eczema). The studies varied in whether they reported diagnostic accuracy of a test for a specific allergen or for all allergens together, and whether they considered accuracy for detecting immediate and/or delayed reactions.

The general trends for sensitivity and specificity of tests for diagnosing any reaction to foods across these studies were as follows:

• The atopy patch test (erythema usually with infiltration) had high (more than 80%) specificity for cow’s milk, egg, soya and peanuts compared with DBPCFC or open food challenges. Specificity results for wheat and sensitivity results for all foods were more variable.
• The skin prick test (wheal size 3 mm or greater) had high sensitivity for egg, fish and peanut compared with DBPCFC; results for cow’s milk, wheat and soya were variable. Sensitivity results compared with open challenge were more variable. Specificity results for all allergens were more variable.
• The sensitivity of specific IgE for cow’s milk and egg was high compared with DBPCFC, but less consistent compared with open food challenge. Specificity results for wheat were more variable. Sensitivity results for all foods were more variable. The specific IgE level indicative of a positive test result was 0.35 ku/l in DBPCFC studies, but ranged from 0.35 to 99 ku/l in the open challenge studies.

Studies that reported the diagnostic accuracy of a test for any food allergen might have been useful for ruling out food allergy, but the available data did not show consistency in sensitivity or specificity results. [EL = DS III]

Outcome of challenge tests

Positive challenges were reported in 40–81% of oral food challenge tests. Egg, cow’s milk and nuts were consistently identified as being the most frequent allergens to trigger an immediate response. Wheat, soya, fish and shellfish were also identified as additional food allergens triggering immediate or delayed responses.

Immediate reactions involved the skin, gut and respiratory systems. Skin reactions were reported as eczematous symptoms or urticaria. Late eczematous reactions occurred in 45% of challenges in two studies. One study reported only a delayed reaction (that is, no immediate response) in 12% of children.

The prevalence of food allergy in children with atopic eczema in secondary care settings was estimated to be 37–56%. [EL = 2]

Management of trigger factors

Most evidence regarding the management of trigger factors in children with atopic eczema related to dietary exclusions or house dust mite avoidance strategies. There was little consistency across studies in the type of diet evaluated, and indications for special diets were not always made clear. There were confounding factors in many studies, for example exclusion of other foods in addition to cow’s milk and egg in studies specifically evaluating exclusion of cow’s milk and egg.

In crossover RCTs, 4–6 weeks’ cow’s milk exclusion diets produced conflicting results with significant differences between treatment and control arms in some, but not all, outcomes. The most common reason for withdrawal from the studies was non-adherence to the diet. [EL = 1−] In infants with moderate to severe eczema and cow’s milk allergy, those fed a whey formula did not exhibit a linear growth pattern during the 9 month follow-up period whereas those fed an amino acid formula did. [EL = 2−]

Egg exclusion alone in children with suspected egg allergy led to improvements in extent, but not severity, of atopic eczema (one RCT). [EL = 1−]

There was no good evidence to support the use of the following interventions in the management of children with atopic eczema: ‘few foods’ diets, elemental diets, addition of probiotics to milk substitutes, sodium cromoglicate (alone or in addition to restricted diets), or excluding foods from the diet of women who are breastfeeding. [EL = 1−]

There was some evidence that house dust mite avoidance strategies in children and adults led to greater improvements in atopic eczema severity than placebo after 2–6 months. [EL = 1−]